[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [idn] Re: Fwd: Unicode letter ballot



On Thu, Nov 28, 2002 at 01:26:16PM +0100, Martin v. L?wis wrote:
> Soobok Lee <lsb@postel.co.kr> writes:
> 
> > > > If option A wins, I see a chaos:
> [...]
> > But, Most homepages in east asia use local charset /encoding. And
> > that requires localcharset<->unicode conversion that produces
> > compatibility CJKs which will be further fed into nameprep.
> > Korean IME produces compatiblity-CJK-equivalent KSX characters in
> > interactive input session on the url bar in browsers.
> 
> Whether this would cause chaos is still questionable. First, I
> understand that these are Chinese compatibility characters, so your
> experience with Korean IME might not be relevant. 

I agree. Korean compatbility  CJKs are in every day CJK use, but CNS11643 
compatibility CJK  may or may not! I guess they will be rarely used.

> The question is
> rather whether users that use IMEs which produce CNS 11643 could enter
> these characters, and whether they could also enter the
> compatibility-normalized character instead.
> 
> [It appears that it will be hard to find an IME which supports the
> full CNS 11643 repertoire in the first place].

Whether or not  full CNS11643 support is rare does nothing to do with
the current issue. In any cases, _if_ someone enters those characters,
he will get errornous results surely! That is the concern.

> 
> In any case, the idea that incorrectly-normalized characters could
> cause chaos seems unlikely.

"Chaos" does not mean stringprep/nameprep is a total failure. For example,
stringprep/nameprep is useful and satisfactory for everyday hangul syllable use!
My concern is that even such partial failures would make chaos _in some
cases_. We can't ignore this. We should do "something", but i am not
sure what to do yet.


> 
> > Yes. This issue is not new problem. We had already discussed . And
> > concluded it is not a big problem. but, this instance of 5 chars
> > seems to need serious decision that we had not expected, iMO.
> > the pure possiblity become the reality.
> 
> What is the change here? Why is U+2F868 more real than U+F951?

As i stated above, whether they are rare/real is not the issue.
Do you agree?

Soobok Lee

> 
> Regards,
> Martin